Evaluate a Personalized Multi Agent System through Social Networks: Web Scraping

Many new applications have been recently developed to satisfy users special needs on the web. In this context, we are interested in personalized systems and particularly in Personalized Multi-Agent Systems (PMAS) characterized by collective and intelligent resolution in a distributed and parallel environment. This work assesses personalization, the most important characteristic of interface in multi-agent systems. As a few studies dealt with the personalization assessment in a multi-agent system, we try, in this work, to address this issue by focusing on web scraping and crawling social networks. In fact, we propose a new assessment tool that exploits data from user's web navigation in order to improve the delivered personalization, which makes the evaluation process more valuable.